Abstract
Applications that deal with high-resolution images, such as virtual reality environments, require visualization devices capable of adequately displaying such images. Actual high-resolution is not possible to acquire without reducing the display's pixel size, and consequently its density per area unit. However, existing technologies do not support the indefinite reduction of the pixel's size. Approaches proposed in the literature to extrapolate the maximum capacity of displays consider characteristics of the human visual system (HVS) and, in most cases, use the composition of multiple visualization devices. One of such approaches consists in the construction of large area display walls obtained through the composition of multiple projectors (called here multi-projector visualization - MPV). Although MPV presents several positive characteristics as, for example, the lack of visible edges when compared to the visualization based on multiple LCD displays, there are still many ways to improve it. In particular, super-resolution reconstruction (SRR) methods can be applied to reach a more expressive resolution enhancement. Currently, sophisticated SRR approaches have applied different techniques and sources of information (such as machine learning, knowledge from the Compressive Sensing framework, and studies based on characteristics of the HVS) to attain such a goal more effectively. In this research project we propose to further explore such possibilities, in the direction of developing new SRR approaches applied to MPV, with the objective of extrapolating the spatial resolution of the used projectors. The project will be executed in a context where the candidate has a large experience - with publications related to SRR in prestigious journals and conference proceedings -, and includes the supervision of a reputed researcher in the area of VMP. It is important to note that the supervisor will make available a richly equipped infrastructure to the researcher, including a laboratory with multiple projectors in different configurations, something that would be impossible to be accomplished with the structure currently available at the candidate's university. (AU)
|